Namboodiri Indian Institute of Technology, Kanpur { badri,sandepkr,vinodkk,vinaypn} @iitk.ac.in Abstract Generating natural questions from an image is a semantic task that requires using visual and language modality to learn multimodal representations. Images can have multiple visual and language contexts that are relevant for generating questions namely places, captions, and tags. In this paper, we propose the use of exemplars for obtaining the relevant context. We obtain this by using a Multimodal Differential Network to produce natural and engaging questions. The generated questions show a remarkable similarity to the natural questions as validated by a human study. Further, we observe that the proposed approach substantially improves over state-of-the-art benchmarks on the quantitative metrics (BLEU, METEOR, ROUGE, and CIDEr). 1 Introduction To understand the progress towards multimedia vision and language understanding, a visual Turing test was proposed by (Geman et al., 2015) that was aimed at visual question answering (Antol et al., 2015). Visual Dialog (Das et al., 2017) is a natural extension for VQA. Current dialog systems as evaluated in (Chattopadhyay et al., 2017) show that when trained between bots, AIAI dialog systems show improvement, but that does not translate to actual improvement for Human-AI dialog. This is because, the questions generated by bots are not natural (humanlike) and therefore does not translate to improved human dialog. Therefore it is imperative that improvement in the quality of questions will enable dialog agents to perform well in human interactions. Further, (Ganju et al., 2017) show that unanswered questions can be used for improving VQA, Image captioning and Object Classification. An interesting line of work in this respect is the work of (Mostafazadeh et al., 2016). Here the authors have proposed the challenging task of generating natural questions for an image. One aspect that is central to a question is the context that is relevant to generate it. As can be seen in Figure 1, an image with a person on a skateboard would result in questions related to the event.

Duane Forrester is the Vice President of Industry Insights for Yext, leading industry outreach, evangelism and authorship for the company. He's here to talk to us about voice search, structured data, artificial intelligence, position zero, integrating your app (or website) into Alexa, and so much more. Duane is is the author of two books: How To Make Money With Your Blog & Turn Clicks Into Customers. He's written for publications ranging from SearchEngineLand and DuctTape Marketing, to Entrepreneur Magazine, the New York Times and Inc. He actively advises startups and large corporations, and even spent time advising the staff who maintain the White House's websites.

Question answering (QA) models have shown rapid progress enabled by the availability of large, high-quality benchmark datasets. Such annotated datasets are difficult and costly to collect, and rarely exist in languages other than English, making training QA systems in other languages challenging. An alternative to building large monolingual training datasets is to develop cross-lingual systems which can transfer to a target language without requiring training data in that language. In order to develop such systems, it is crucial to invest in high quality multilingual evaluation benchmarks to measure progress. We present MLQA, a multi-way aligned extractive QA evaluation benchmark intended to spur research in this area. MLQA contains QA instances in 7 languages, namely English, Arabic, German, Spanish, Hindi, Vietnamese and Simplified Chinese. It consists of over 12K QA instances in English and 5K in each other language, with each QA instance being parallel between 4 languages on average. MLQA is built using a novel alignment context strategy on Wikipedia articles, and serves as a cross-lingual extension to existing extractive QA datasets. We evaluate current state-of-the-art cross-lingual representations on MLQA, and also provide machine-translation-based baselines. In all cases, transfer results are shown to be significantly behind training-language performance.

Machine Learning in Pharmaceutical Market Research Report has been studied and presents an actionable idea to key contributors working in it. A thorough study of the competitive landscape of the global Machine Learning in Pharmaceutical Market has been given, presenting insights into the company profiles, financial status, recent developments, mergers and acquisitions, and the SWOT analysis. This report has published stating that the Global Machine Learning in Pharmaceutical Market is anticipated to expand significantly at Million US$ in 2019 and is projected to reach Million US$ by 2026, at a CAGR of during the forecast period. The global Machine Learning in Pharmaceutical market can be segmented based on product type, application, end-user, and region. This report gives an in depth and broad understanding of market with accurate data covering all key features of the prevailing market, this report offers prevailing data of leading companies.

Australian IBM cybersecurity engineers have developed an artificial intelligence (AI) system to analyse network connections and employee communications at an enterprise scale. The model detects changes in users' behaviour and can automatically triggers investigations even if the changes occur across multiple platforms. IBM research found the root cause for 52 per cent of data breaches in Australia was malicious or criminal attacks which often use methods like phishing and social engineering. The new IBM solution, developed in the company's Gold Coast cybersecurity lab as part of a hackathon, uses AI to monitor changes in employee behaviour and flags indicators of compromise. It was debuted to the industry at last week's Australian Cyber Conference in Melbourne as a way of showing what can be done but the solution is not something that can be bought directly from IBM. Currently known as "QRadar Insider Threat Detector with Watson" it uses IBM's AI model, Watson, to analyse user generated content – like emails, Word documents, and Slack messages – to detect both the tone of content and employees' typical behaviour or "personalities".

In 2018, voice search was one of the hottest topics in the SEO community. A popular article by Wordstream listed a handful of statistics around voice search, starting with the misconstrued Comscore statistic that by 2020, 50% of searches would be done through voice. It turns out, this statistic was related only to voice search in China. Despite the inaccuracy in the U.S. and overall global market, the quote has reverberated through the SEO industry and pushed digital marketers to frantically prepare themselves by learning everything they could about voice search optimization. As 2020 approaches, marketers are now skeptical voice search will actually cause a cataclysmic shift to our marketing strategies.

I look forward to the next two posts. I hope you cover the challenges with public AI apps. Being able to ensure data privacy is maintained by being able to use this technology in private clouds or on premise would be helpful. Application portability and data security is also key.

AI has gone through many cycles since we first coined the term "machine learning" in 1959. Our latest resurgence began in 2011 when we put Watson on national television to play Jeopardy! This became a cornerstone event, demonstrating that we had something unique. And we saw early success, putting Watson to work on projects with clients. This created even more excitement. That excitement led to more opportunity.

Multi-hop question answering requires models to gather information from different parts of a text to answer a question. Most current approaches learn to address this task in an end-to-end way with neural networks, without maintaining an explicit representation of the reasoning process. We propose a method to extract a discrete reasoning chain over the text, which consists of a series of sentences leading to the answer. We then feed the extracted chains to a BERT -based QA model (Devlin et al., 2018) to do final answer prediction. Critically, we do not rely on gold annotated chains or "supporting facts:" at training time, we derive pseudogold reasoning chains using heuristics based on named entity recognition and coreference resolution. Nor do we rely on these annotations at test time, as our model learns to extract chains from raw text alone. We test our approach on two recently proposed large multi-hop question answering datasets: WikiHop (Welbl et al., 2018) and HotpotQA (Y ang et al., 2018), and achieve state-of-art performance on WikiHop and strong performance on HotpotQA. Our analysis shows properties of chains that are crucial for high performance: in particular, modeling extraction sequentially is important, as is dealing with each candidate sentence in a context-aware way. Furthermore, human evaluation shows that our extracted chains allow humans to give answers with high confidence, indicating that these are a strong intermediate abstraction for this task. 1 Introduction As high performance has been achieved in simple question answering settings (Rajpurkar et al., 2016), work on question answering has increasingly gravitated towards questions that require more complex reasoning to solve. Multi-hop question answering datasets explicitly require aggregating clues from different parts of some given documents (Dua et al., 2019; Welbl et al., 2018; Y ang et al., 2018; Jansen et al., 2018; Khashabi et al., 2018).